Poster Talks Explore How Policy, Payer Incentives Affect Cancer Care, Outcomes

For all the therapeutic advances presented during the 2023 American Society of Clinical Oncology (ASCO) Annual Meeting, there are simple realities: patients cannot be treated if they have no access to care. Too many patients miss a chance at survival for such reasons: they lack health coverage, they can’t take a day off, or they can’t navigate the system.

At the same time, therapy given when patients won’t benefit helps no one; this drives up the costs and wastes resources better spent improving the quality of end-of-life care.

No one knows this more than oncologists themselves, and on June 5 a group gathered for a session of talks on care delivery and regulatory policy.

Blau

Williams

This session, chaired by Sibel Blau, MD, medical director of Northwest Medical Specialties and president and CEO of the Quality Cancer Care Alliance, and Vonetta Michele Williams, MD, PhD, radiation oncologist at Memorial Sloan Kettering Cancer Center, invited speakers to frame presentations around 3 related posters, selected from the thousands accepted for the ASCO meeting. The first 2 talks addressed payer resource issues, including the now-expired Medicare payment model, the Oncology Care Model (OCM).

Medicaid Expansion and Paid Leave Both Linked to Improved Survival

Matthew F. Hudson, PhD, MPH, associate professor of medicine at the University of South Carolina School of Medicine, Greenville, discussed 3 posters involving access to care:

• An analysis led by Xuesong Han, PhD, director of Surveillance and Health Services Research at the American Cancer Society, compared data for young women with breast cancer in Medicaid expansion states and those that did not expand Medicaid; they found that Medicaid expansion was linked to increases in guideline-concordant care, fewer treatment delays, and improved 2-year survival. More shocking was this finding: guideline-concordant care and 2-year survival rates for young patients with breast cancer fell over time if states failed to expand Medicaid.¹
• Samuel U. Takvorian, MD, MS, assistant professor of medicine, Hospital of the University of Pennsylvania, and colleagues found that Medicaid expansion was associated with a significant increase in the share of Black and Hispanic patients taking part in oncology clinical trials if states required Medicaid to cover trial costs.²
• A study led by Justin Michael Barnes, MD, MS, radiation oncologist, Washington University St. Louis, found state paid medical leave requirements were linked to improved survival for working-age adults with newly diagnosed cancer, especially if they lacked insurance.³

Why don’t such findings lead to better access for all? As Hudson explained, developing these data is just the first step. ‘We can glean from this that federal and state policy favorably correlates with evidence generation, cancer care, delivery, research, and health outcomes,” he said.

Hudson

“But at this point, we might need a reminder fundamentally to explore, what, in fact, is policy?”

Policy, he asserted, has many elements: it is a set of principles, it requires people to act, and it calls for assessments of those actions. The challenge with implementing policy, Hudson said, is that it requires a level of consensus that may or may not exist:

“If we contemplate policy as a principle or tool, we’d argue that state policy provides knowledge,” he said. “That is, it publicly declares an ethical stance of right and wrong. It also encourages or challenges moral perceptions, the subjective beliefs related to this ethical stance.”

Hudson continued: “It assumes skills. That is, it assumes a compliance capacity, and it expects compliance behavior. Now, it expects this of individuals of groups and organizations, all in service of advancing patient outcomes. So, policy as an intervention expects [that] its mere presence begets organization, group, and individual change that produces patient-centered outcomes.”

The results Hudson highlighted—which showed both the benefits of access to care and the harms of poor access—point to the next phase of research, which would be policy implementation science. Who are the “actors” at the state or federal levels who can advance the public’s health? What steps would be needed to translate these results into action?

While Han and colleagues focus on health care access, Hudson said, “I think there's an opportunity in future research to think critically about which regional planning policies interact with Medicaid policy to optimize breast cancer outcomes. We may also consider how regional planning policies drive treatment facility availability and its characteristics.”

The opportunity, Hudson said, is “to consider whether treatment affordability impacts survival through adherence or treatment choice.”

He also encouraged the oncologists to start weighing the impact of “time toxicity,” a new concept emerging in cancer care, which considers the time patients spend dealing with the health system bureaucracy and how that affects survival. Simply put, patients who can’t take time off or have no one to help them with prior authorization may give up and forgo treatment.

“How might patients trade home days for survival days?” Hudson asked. “And how might that mediate the impact of paid medical leave policy?”

Evaluating Success in Care Coordination for Patients With Cancer

Adelson

Kerin Adelson, MD, chief quality and value officer at The University of Texas MD Anderson Cancer Center, next discussed research on a group of initiatives designed to improve patient care, and, where possible, reduce costly and futile end-of-life interventions.

Community health workers. A clinical trial led by Manali Patel, MD, MPH, MS, Engagement of Patients with Advanced Cancer (NCT02966509) randomized 213 veterans with stage III or IV cancer to usual care (UC) or usual care plus a 6-month intervention with a community health worker (CHW), who helped with advanced care planning, completion of advance directives, and discussions of goals and values with their clinicians.⁴

Although prior work has shown how such efforts promote discussions and shrink acute care costs, Adelson said, “Dr Patel asks what the long-term effects of the community health worker intervention were on the overall survival and health care use in the last 30 days of life.”

After a median follow-up of 298 days, results were significant:

• The CHW group had improved overall survival compared with the UC group, with a hazard ratio (HR) of 0.75 (95% CI: 0.56-0.98), P = .04.
• There CHW group used less than half the amount of emergency department (ED) and hospital stays in the last 30 days of life as the UC group, with odds ratios (OR) of 0.33 (0.18-0.63) for ED use and 0.28 (0.15-0.54) for the hospital.
• The CHW group was also more likely to use palliative care than the UC group (OR 1.86; 1.03-3.33) and hospice (OR 2.36; 1.28-4.36).

Adelson showed the Kaplan-Meier curves for survival results and remarked, “You could imagine a phase 2 trialist for a therapeutic drug just being ecstatic to show this kind of result.”

“What is it about a community health worker intervention that worked so much better than many prior studies?” she asked. “Is there more trust when the conversations involve a non- medical person from the patient's own community? I would love to know what actually happened in the interactions between the community health worker and the patients,” she said, suggesting a qualitative analysis.

“Finally,” Adelson asked, “Why did the intervention group live longer? Was it related to the higher rates of palliative care use? Was it that help with coping that gave them more reason to live?

Oncology Care Model (OCM). Adelson next took up a timely abstract that examined the OCM, Medicare’s 6-year alternative payment model to improve cancer care. The OCM ended a year ago (and is to be replaced with the Enhancing Oncology Model as Evidence-Based Oncology goes to press). The analysis, led by Gabriel A. Brooks, MD, MPH, an associate professor of medicine and The Dartmouth Institute at Dartmouth Geisel School of Medicine, reviewed data through year 5 of the program.⁵

The authors concluded that the OCM led to a $499 relative reduction in Medicare payments for 6-month episodes, with similar care quality; the biggest savings impact was seen in high-cost cancers; the OCM had no overall impact on chemotherapy spending, but reduced spending on supportive care; and the OCM had no overall impact on ED visits or hospitalizations, relative to practices outside the OCM.

The authors concluded that a significant reason Medicare lost money on the OCM was the size of payments for monthly enhanced oncology services (MEOS), which covered required items such as patient navigation, advanced care planning, and 24/7 access to records. A major source of controversy among practices considering the EOM is the cut in MEOS payments from $160 per patient per month (PMPM) under the OCM to $70 PMPM.

Also, as practices became more nimble with the OCM, they were also more successful in earning back savings as performance payments, which further eroded Medicare’s savings.

Adelson reviewed which items were and were not included in the OCM quality measures, noting that these included things such as curtailing ED and hospital visits, increasing use of hospice, and screening for pain and depression. Although practices were required to use clinical guidelines, they did not have to demonstrate use or calculate use of systemic therapy at end of life as quality measures. More importantly, Adelson shared an analysis from Tuple Health based on data from Yale Medicine, which showed how drug costs had overtaken other aspects of cancer care: even if oncology teams found savings elsewhere, it was hard to contain overall costs when drugs climbed from 53% of total cost in 2012 to 71% by 2020.

“In the age of molecular oncology, there are really very few examples of clinical equipoise between high cost and low cost, disease modifying therapies. And I actually think it's a fantastic sign that the practices demonstrated a willingness to make substitutions when they felt they were clinically appropriate,” she said.

In her conversations with colleagues, Adelson hears of great willingness to substitute biosimilars or other less expensive therapies. “But again, there are very limited opportunities.”

While noting the authors’ findings, Adelson said Medicare lost money because of the program design, not from lack of effort by the practices. “Whether the program is successful or not, it's a matter of perspective,” she said. “If you wanted oncology practices to improve quality, at least to some extent, and reduce Medicare billing, then the program was absolutely a success. If you were trying to save Medicare money—if that was your goal, then the program design was flawed from the outset.”

Instead, she said, a payment model more like ASCO’s Oncology Medical Home—which addresses clinical nuances and evaluates practices for values within their control—would do more to promote good care.

Offering a nudge. Takvorian was the lead author for a second abstract, which reported results of a clinical trial (NCT04867850) on the effect of electronic “nudges” to both clinicians and patients to promote

Takvorian

conversations about serious illness. Between September 2021 and March 2022, the study tracked 4450 patients seen by 166 clinicians. Evaluations were randomized to the clinician nudge (1179), the patient nudge (997), both (1270) or the active control (1004).

Compared with the control, serious illness conversations were more likely among participants in the combination nudge arm, with an HR of 1.55 (95% CI, 1.00-2.40); the conversations were not more likely in the clinician arm (HR 0.95; CI 95%, 0.64-1.41) or the patient arm (HR 0.99; 95% CI, 0.73-1.33) relative to the control. The authors concluded that a system of 2-way nudges may have a “synergistic” effect in encouraging serious illness conversations in an equitable way across a large health system.⁶

Adelson reviewed the process that the Penn Medicine team followed for the nudges, including an element of competition among the oncologists in one arm by offering comparative feedback.

Although the increase in conversations at 6 months for the combined nudge group was statistically significant, Adelson cautioned, “It was still only 14% of patients who are actually having these conversations, which says we still need to do more as a society. And we need to look at studies like Dr Patel's that were so successful in moving this needle.”

Misaligned incentives. In looking at which interventions improve survival, Adelson said it’s also important to look at what does not help survival: therapy at end of life.

Then, she shared data her team had presented at ASCO. “One of the criticisms of the of the metric looking at chemotherapy within 14 days of death is that we're missing the patients treated at the same time point who go on to live and have the benefit,” she said. “If that’s true, when we compare practices that give lots of chemotherapy at the end of life to those that give less, we should see improved survival for the more aggressive practices.”

Adelson and her team performed this analysis using Flatiron’s database, examining the 6 most common cancer types, looking at the entire population and not just those who died.⁷ “There is absolutely no survival benefit for more aggressive use of chemotherapy in the end-of-life setting,” she said. And, this affects downstream care, too, with higher rates in the intensive care unit, higher hospital admissions, and lower hospice rates.

“So, maybe we have some misaligned incentives,” Adelson said. “We know that these patient centered interventions improve survival, reduce acute care use, improved patient and caregiver experience, reduce cost, and for the most part are not reimbursed,” except perhaps under E/M billing for palliative care visits.

“Conversely, aggressive use of end-of-life systemic therapy has no survival benefit, increases acute care use at the end of life harms patient and caregiver experience, increases cost and could be reimbursed as high as $80,000 a month,” she said, using a common immunotherapy as an example. “The money saved on end of life oncolytics and downstream acute care could be redirected to enhanced home hospice services, more flexible criteria for inpatient hospice, and communication training programs.”

References

1. Han X, Shi KS, Ruddy KJ et al. Association of Medicaid expansion with treatment receipt, delays in treatment initiation, and survival among young adult women with breast cancer. J Clin Oncol. 2023;41(suppl 16):abstr 1511. 10.1200/JCO.2023.41.16_suppl.1509

2. Takvorian SU, Schpero WL, Blickstein D, et al. Association between state Medicaid policies and accrual of Black and Hispanic patients to cancer clinical trials. J Clin Oncol. 2023;41(suppl 16):abstr 1511. 10.1200/JCO.2023.41.16_suppl.1510

3. Barnes JM, Chino F, Johnson KJ, et al. State mandatory paid medical leave policies and survival among adults with cancer in the US. J Clin Oncol. 2023;41(suppl 16):abstr 1511. 10.1200/JCO.2023.41.16_suppl.1511

4. Patel MI, Agrawal M, Kapphahn K, et al. Engagement of patients with advanced cancer (EPAC) randomized clinical trial: long-term effects on survival and health care use. J Clin Oncol. 2023;41(suppl 16):abstr 1513. 10.1200/JCO.2023.41.16_suppl.1513

5. Brooks GA, Trombley M, Landrum MB, et al. Impact of the Oncology Care Model (OCM) on Medicare payments, utilization, and care delivery: update through year 5. J Clin Oncol. 2023;41(suppl 16):abstr 1512. 10.1200/JCO.2023.41.16_suppl.1512

6. Takvorian SU, Clifton ABW, Gabriel PE, et al. Patient- and clinician-directed implementation strategies to improve serious illness communication for high-risk patients with cancer: A cluster-randomized pragmatic trial. . J Clin Oncol. 2023;41(suppl 16):abstr 1514. 10.1200/JCO.2023.41.16_suppl.1514

7. Adelson K, Canavan M, Sheth K, Scott JA, Westvold S. The impact of receipt of systemic anticancer therapy (SACT) near the end of life (EOL) on cost among oncology practices participating in CMS' Oncology Care Model (OCM). J Clin Oncol. 2023;41(suppl 16):abstr e18923. doi: 10.1200/JCO.2023.41.16_suppl.e18923